The use of artificial intelligence in marine research will open the door to new possibilities
About three-quarters of the earth's surface is the sea. As a result, the oceans still have great potential for everything from climate control to the transportation of goods around the world, use as a source of clean energy, research and extraction of polymetallic nodules from minerals.
A polymetallic nodule is a type of mineral that contains the 4 most important components of a battery: cobalt, nickel, copper, and manganese.
In other words, many potential areas of the sea have not yet been exploited to our advantage. The United Nations has declared the current decade from 2021 to 2030 as the Decade of Marine Science for Sustainable Development.
The announcement is aimed at uniting countries that have global maritime boundaries to reduce pollution and stabilize the oceans.
Creating possibilities by increasing knowledge about the sea
A recent study by the University of Bath in England used two AI algorithms to understand the echoes of sonar in the ocean. The ‘sonar’ device or method is usually used to detect and locate an object submerged in water with the help of reflected sound waves. The aim of this study, however, was to observe the changes in the echo of gold in the depth, salinity and temperature differences of the ocean through AI.
The algorithms were able to classify the environment of different regions of the seabed with more than 90% accuracy from ‘gold measurement’.
Combining technologies such as artificial intelligence, machine learning algorithms and smart robots, it is possible to go a long way in marine research in the near future. It was difficult even a few years ago to conduct deep sea excavation or research on the deep sea without disrupting the life of animals or plants living under the sea. But now it is not as difficult as before.
With the help of these modern technologies, oceanographers are able to create accurate maps of the sea floor. At the same time the effects of climate change, the latest status of different species of animals or the salinity levels of different regions of the sea can be easily determined. Also not yet discovered, they are also able to gather a lot of information about such areas of the sea.
Research on the life of sea creatures seems to be quite exciting. But actually collecting and processing information under the sea is quite difficult. To address this, the Marine Robotics Innovation Center has partnered with a number of industry partners at the National Oceanography Center in the United Kingdom to develop an automated vehicle that can run under the sea. The name of the vehicle is Autosub 2KUI. The vehicle is equipped with sophisticated gold instruments and camera system. As a result this go can work even under ice.
By analyzing the data thus collected through deep-learning algorithms, it is possible to infer patterns related to deep-sea life.
Accurate mapping of ocean spaces and accurate information on their location and quantity are needed to extract resources. 3D visual imaging technology, equipped with a biocam or high-sensitivity detector, can be used to capture deep-sea fauna, volcanic eruptions and seabed surface color. These images clearly show the effects of rising sea temperatures and acid levels on the lives of various microorganisms, including corals. These images will also be used to create the necessary database for conducting excavations at sea.
It is a difficult task to study the classification of a particular species as it is home to many different species of microorganisms with similar characteristics. But now it is possible to solve this problem through Convolutional Neural Networks (CNN's). This technology is a weight-sharing network based on photo classification, which is being used to classify different images of the seabed.
Processing images has become much easier due to the skill in computing and the ability of this technology to learn on its own. The two most important parts of this technology are the convolution kernel and the output. The Convolution kernel allows you to identify and test various features of an image. And through the output that picture is classified.
In addition, some more picture-based AI technologies can help us understand different parts of the observed microorganisms. This makes it easier to classify them. Machine learning can use a variety of algorithms to make connections between plankton images at sea and the genomic data of those animals. As a result, various genetic information responsible for their size can be identified and examined.
To advance research on the sea, it is necessary to build specialized new ports, build infrastructure for power generation such as off-shore wind farms, and open new shipping lanes. However, sound can travel through water at least 4 to 5 times faster than air. Noise generated from any construction work or container movement is also responsible for ‘noise pollution’ in water. But this is rarely discussed.
This noise pollution affects the livelihoods of several marine animals, including whales and dolphins. Because these animals are very sensitive to high levels of noise. With the help of machine learning technology, the marine animals in the vicinity can be identified instantly. And it is possible to control the sound by maintaining balance on the basis of the data obtained through it.
With the development of artificial intelligence, robotics and machine learning technology, various missions can be conducted at sea without harming marine animals or the environment. With the help of these technologies it will be possible to create awareness about the sea in addition to collecting information on many unknown things. As a result, we will be able to improve our activities to stop pollution at sea before it is too late.
Tags:
TIPS